Jyunmi Hatcher

But if we can speed up at least this part of it, we can get more solutions for more people faster and help more people.

So why is this a bigger story than singly one study?

So first, democratization.

I'm a big fan of lowering barriers.

I would love it if models got good enough and tight enough to where I can run a full-size model on my laptop or even better, on my phone.

And if everyone could run it on their own hardware, that'd be great.

So because a high school student just co-authored a paper in Cell Reports Medicine, the barrier for doing serious data science just dropped dramatically.

The skill is shifting from can you write the code to can you ask the right question?

And can you ask the right question seems to be...

one of the main things that I'm seeing collectively in the AI space.

Can you write the right question?

Can you provide the ideal example?

So if you have these kinds of skills, that's going to translate to great insights and output from AI in general.

Just like months ago, we were talking about having that ability to constantly want to learn and having agency to just do it.

Doesn't have to be perfect.

You can ideate, right?

Second, like I said, speed saves lives.

In preterm birth research, there are thousands of babies born too early every day.

Compressing a two-year research cycle to six months has real consequences for how fast new tools reach patients and doctors.

And third...